Ventilatory responses to hypocapnic vertebral artery perfusion in intact and carotid body denervated dogs

Re.spiration Physiology (1981) 45, 97-110 Elsevier/North-Holland Biomedical Press

VENTILATORY RESPONSES TO HYPOCAPNIC VERTEBRAL ARTERY PERFUSION IN INTACT AND CAROTID BODY DENERVATED DOGS*

PAUL J. FEUSTEL**, J. MILTON ADAMS, DAVID F. DONNELLY and ROBERT E. DUTTON Department of Physiology, Albany Medical College of Union University, Albany, N Y, U.S.A.

Abstract. The ventilatory responses to step changes in vertebral artery Pco2 were investigated in intact and carotid body denervated dogs. The steady-state ventilatory responses of the denervated dogs were less than those of intact dogs. However, when expressed as a ratio to the control ventilation there was no difference between the two groups. While the arterial Pco2 was held at 56 m m Hg by adding CO 2 to the inspired air the perfusion of the vertebral arteries was switched from the dog's own arterial supply to hypocapnic blood. The ventilation of the denervated dogs decreased at a faster rate (half time = 130 +_ 9 sec) to a level less than the r o o m air control ventilation. The ventilation in the intact dogs decreased at a slower rate (half time = 184 _+ 23 sec) and was maintained above the room air control level after ten minutes o f hypocapnic perfusion. Increasing the medullary blood flow, as measured with radiolabeled microspheres, augmented the rate of decline of ventilation in intact dogs. We conclude, (1) the influence of the peripheral chemoreceptors appears to increase as central drive is decreasing, and (2) the remaining time course of the decrease in ventilation is related to the rate of brain stem perfusion. Cerebral blood flow Chemoreceptors Control of respiration

Hypocapnia Ventilatory transients

Although the interaction of the central and peripheral chemoreceptors has been extensively studied during transient stimulation of either the peripheral chemoreceptors alone (Edelman et al., 1973; Dutton et al., 1964) or both chemoreceptors simultaneously (Gelfand and Lambertsen, 1973), the effect of transient removal of stimulation from the central chemoreceptors has not been widely studied. The AcceptedJbr publication 28 March 1981 * This research supported by N I H HL-12564 and HL-07194. ** Present address: Anesthesia Research Center, Cardiovascular Research Institute, R o o m 1386-HSE, University of California School of Medicine, San Francisco, CA 94143, U.S.A. 0034-5687/81/0000b0000/$ 02.50 © Elsevier/North-Holland Biomedical Press 97

98

P.J. FEUSTELet al.

vertebral artery perfusion technique has been employed to test the interaction during transients of CO 2 or pH applied centrally while a steady level of COg is maintained peripherally (Fitzgerald, Gross and Dutton, 1968; Dutton et al., 1969). In dogs with intact chemoreceptors, only a 33% reduction in ventilation was observed following two minutes of hypocapnic perfusion of the vertebral arteries superimposed on a systemic hypercapnia (Dutton et al., 1969). In order to determine whether the peripheral chemoreceptors are wholly or in part responsible for the sustained ventilation under these conditions this procedure was repeated in intact and carotid body denervated dogs. Since central chemoreceptors are believed to be located approximately 200 microns below the ventrolateral surface of the medulla and are responsive to changes in cerebrospinal fluid Pco~ and pH (Mitchell et al., 1963), another mechanism by which ventilation may be maintained during hypocapnic perfusion is that CO 2 in CSF and surrounding tissue serves as a reservoir to delay the changes in Pco~ in the receptors. Also, since arterial hypocapnia rapidly reduces cerebral blood flow (Severinghaus and Lassen, 1967), low levels of blood flow might further prolong the time required for the removal of the central CO2 stimulus. To determine the extent to which this mechanism contributes, perfusions were also done at higher flow rates. The vertebral artery contribution to medullary blood flow was measured with microspheres. A third possible mechanism for the maintained ventilation is neural reverberating circuits (Salmoiraghi and Burns, 1960) which, having been excited by increased levels of carotid body afferent activity during hypercapnia, can decay only slowly, despite the removal of CO2 stimulation. Such a phenomenon has been observed following chemical stimulation of the carotid body with CO 2 (Dutton et al., 1967) as well as after hyperventilation induced by electrical stimulation of the carotid sinus nerve (Eldridge, 1973). The high flow perfusions serve to further define this phenomenon. If this mechanism is independent of the level of central chemoreceptor activity then increasing the rate of removal CO 2 stimulation should have no effect on the time course of recovery.

Methods

Carotid body denervations were performed under sodium pentobarbital anesthesia (25-30 mg/kg) using a technique described by Heymans and Bouckaert (1934) and modified by Krasney (1971). At least a week prior to the perfusion experiment both carotid sinus nerves were ligated and sectioned. Denervations were tested twice by injections of a 2-ml bolus of 3 mM sodium cyanide directly into the common carotid artery, once immediately following denervation and a second time at the conclusion of the vertebral perfusion experiment. Any changes in respiratory rate or depth was evidence of incomplete denervation and either a second attempt was made to complete the denervation or the experiment was discarded. During the denervation

HYPOCAPNIC VERTEBRALARTERY PERFUSION

99

procedure the occipital arteries, which anastomose between the vertebral and the carotid arteries were ligated. In intact dogs this ligation was done just prior to the vertebral perfusion, in order to prevent hypocapnic blood from reaching the carotid bodies. At the time of the vertebral perfusion experiment, dogs weighing between 20 and 25 kg were anesthetized with sodium pentobarbital (25-30 mg/kg). Catheterizations of both femoral arteries and one femoral vein were performed. Systemic arterial pressure was monitored from one femoral artery by a Statham P23Db strain gauge and an Electronics for Medicine recorder. The remaining arterial cannula was used for withdrawal of blood during the perfusion so that a constant arterial pressure was maintained. Systemic arterial blood pressure was maintained above 80 mm Hg throughout the experiment. The femoral venous cannula was used for supplementary anesthesia and anticoagulant administration. Infusate blood was obtained from a donor dog, exchanged with the experimental dog and placed in a temperature and humidity controlled tonometer. Blood was equilibrated to a Pco.~ of less than 10 mm Hg and a Po2 matched to a previous arterial blood sample. Blood gas analysis for Pco2, Po2 and pH were obtained by means of Severinghaus, modified Clark and capillary glass electrodes, respectively. Values were corrected for temperature. A tracheostomy was performed and the tracheal cannula was connected to a Fleisch pneumotachograph and a Statham PMS differential pressure transducer. The output of the transducer was amplified, gated and integrated to give expired tidal volumes. Pco2 was continuously monitored with a Godart infrared CO2 analyzer. CO2 was added to the inspired air and the flow was continuously adjusted to maintain the end-tidal CO 2 at eight percent. All data were recorded on a Honeywell 5600 FM tape recorder and subsequently analyzed by digital computer. Each breath was assigned to a ten second window and the average for that window calculated for each experiment. The vertebral perfusion technique has been previously described (Fitzgerald et al., 1968 ; Dutton et al., 1969). Cannulation of the vertebral arteries was done proximal to anastomosis between the vertebral and the anterior spinal arteries. Prior to the insertion of the vertebral artery loops, the dog was heparinized (5000 units). Supplementary heparin (1000 units) was given every hour. The distal end of each cannula was inserted into the vertebral arteries. The proximal ends were placed into the ipsilateral brachial arteries. Two side branches on each loop enabled monitoring of pressure and the perfusion of the vertebral arteries with equilibrated blood from an external perfusion system. A Biotronics cannulating electromagnetic flow probe which had been soaked for 24 hours in Tridodecylmethyl ammonium chloride-heparin complex (2~o solution) was placed in each loop. Clamps were placed on the tubing at appropriate points so that perfusion could be done from either the dogs own arterial system or from the perfusion system. Each experimental sequence consisted of a 2-min air breathing control period, a 5-min CO 2 breathing period during which the end-tidal fraction of CO 2 was held

100

P.J. FEUSTEL

et al.

at 8~o by adjusting inspired CO 2, and a perfusion period during which hypocapnic blood was infused into the vertebral arteries while the end-tidal CO 2 was continuously held at 8 ~ . Perfusion patterns tested were (1) constant pressure, (2) constant flow, and (3) high flow perfusions. A servo controlled Harvard Apparatus peristaltic pump held vertebral artery pressure 5 mm Hg above mean arterial pressure for the constant pressure perfusions. During the constant flow perfusions the pump was set to maintain the flow equal to that measured during the CO2 breathing period. Since no difference was found between the ventilatory responses of the constant pressure and the constant flow perfusions, these results were combined. A high level of perfusion was achieved by setting the pump to deliver a flow approximately twice that measured during the CO 2 breathing period. The constant pressure and constant flow perfusions were continued for 10 min or until the tidal volume decreased to one third of the air breathing tidal volume. This latter constraint was necessary to avoid the irreversible damage from systemic hypoxia which occurred in preliminary expriments on denervated dogs when they became apneic. High flow perfusions were of shorter duration because the supply of perfusate was exhausted within 3 to 5 min. Control perfusions were done to insure that the perfusion itself did not affect ventilation. The perfusate consisted of blood drawn slowly from the vertebral arteries during the CO2 breathing period. The perfusions were then done in the same way as the hypocapnic perfusions. At the conclusion of the experiment an infusion of 0.2 to 1.0 ml of 107/o ethanol in saline was used to determine the transit time from the perfusion tubing to the medulla (Crone, 1965). The end point of the ethanol test was a transient apnea for an interval equivalent to two or three breaths and a transient decrease in blood pressure. This time was subsequently subtracted from the record of ventilation for constant pressure and constant flow perfusions. In no instance did the ventilation change faster than the transit time for ethanol. Verification and quantification of medullary flow were obtained by multiple injections of radioactive 15 micron microspheres (3M Corporation). Three varieties of microspheres were used (~Sstrontium, 5~chromium, and ~4~cerium). They were injected into the vertebral artery perfusate via a mixing chamber in 9 intact dogs during a CO~ breathing period, following 3 min of a normal flow hypocapnic perfusion, and following 2 min of a high flow hypocapnic perfusion. The dogs were killed, the brains removed, fixed in 3%0 formalin, sectioned, weighed and the radioactivity determined with a Nuclear Chicago gamma counter. After correction for isotope decay and overlap of isotope energies by a technique similar to that of Rudolph and Heymann (1967), the flows were determined on a per gram basis. By determining the percentage of injected counts in each section, the fraction of total vertebral artery flow perfusing each section was determined. The total vertebral artery flow was measured at the time of microsphere injection either by the electromagnetic flow probes in the case of CO2 breathing determinations or by the pump speed in the case of perfusion determinations. Unless otherwise indicated all statistical analysis was done with Students ~t'-test with a significance level of 0.05.

HYPOCAPN1C VERTEBRAL ARTERY PERFUS1ON

101

Results A total of 40 steady-state ventilatory responses to CO2 were measured in 12 peripheral chemoreceptor intact dogs and 33 responses were determined in six peripheral chemoreceptor denervated dogs (table 1). Although the steady-state ventilation during both the air breathing and the CO,, breathing period are significantly greater in the chemointact dogs, the percentage increases in ventilation were not different. Ventilation increased by 374 +_27~0 in chemointact dogs and by 414_+ 51~o in chemodenervated dogs. These increases in ventilation resulted from increases in both frequency and tidal volume in both chemointact and chemodenervated dogs. Arterial and infusate blood gas values are shown in table 2. In chemointact dogs, ventilation at the end of the 20 normal flow hypocapnic perfusions was 13.4+ 2.2 1/min and exceeded the preceding r o o m air control ventilation (paired 't'-test, P < 0.05). The final level of ventilation in the chemodenervated dogs was 5.3 _+ 1.9 l/min and was less than the corresponding r o o m air control ventilation (paired 't'-test, P < 0.05). O f the 33 hypocapnic normal flow perfusions in chemodenervated dogs, four perfusions were terminated between three and five minutes and an additional nine were terminated between five and ten minutes due to low tidal volumes. Perfusions were eliminated from the ten second averages shown in fig. 1 and all subsequent figures when their tidal volumes dropped below one half of the r o o m air control tidal volume. Therefore, the fact that the mean ventilation dropped to a final value less than r o o m air control does not appear in the figures. After 20 sec of normal flow hypocapnic perfusion, ventilation, when expressed as a ventilation ratio (VR) to the 8% end-tidal CO: ventilation, dropped significantly in both the chemointact and chemodenervated dogs. There was no difference

TABLE 1 Steady-state ventilatory data for chemointact and chemodenervated dogs (Values are mean ± standard error of mean)

Chemointact (N = 20) Room air control 8% End tidal Chemodenervated (N = 33) Room air control 8% End tidal

Ventilation (L/min)

Tidal volume (ml)

Frequency (/min)

8.2 ±0.8 39.3 ±3.3

265 ±55 625 ±81

29.5 ±1.7 54.3 ±2.0

5.6 ±0.5 26.7 ±2.3

165 ±13 489 ±39

35.6 ±1.8 54.2 ±2.0

102

P.J. F E U S T E L et al.

TABLE 2 Arterial and infusate blood gas data for chemointact and chemodenervated dogs (Mean _+ SEM)

Chemointact (N = 20) R o o m air control 8~o End tidal CO2 Infusate

Chemodenervated (N = 33) R o o m air control 8 ~ End tidal CO 2 Infusate

PCO2 (Tort')

PO2 (Torr)

pH

30.9 _+ 1.7 55.5 _+ 1.1 7.4 _+0.4

86.0 _+ 1.9 108.9 _+ 1.8 112.0 _+4.0

7.354 _+0.008 7.181 _+0.010 7.644 _+0.020

30.7 _+ 1.2 57.1 _+1.2 9.9 _+0.8

80.0 _+3.4 97.3 _+3.1 81.1 _+3.1

7.332 _+0.017 7.179 _+0.021 7.584 _+0.021

I.oi ~ QI ITI

0.6

0% .... TIME (seconds) Fig. 1. The transient ventilatory responses (means _+ SEM) of chemointact and chemodenervated dogs to hypocapnic vertebral artery perfusion at normal flows. Ventilation is expressed as a ratio (VR) to the CO 2 ventilation. Perfusion began at t = 0. R o o m air VR was 0.21 + 0.02 for the chemointact and 0.28 _+ 0.02 for the chemodenervated dogs.

between the ventilatory responses of the chemointact and chemodenervated dogs for the first 70 sec, however at 80 sec the V R of the chemointact dogs was 0.82 _+ 0.04 while in the chemodenervated dogs it was 0.73 _+ 0.02. The difference in responses, which was significant at 80 sec, continued and increased during the remainder of the perfusions. The mean half times for the ventilatory responses during hypocapnic normal flow perfusions were 184_+23 sec in the chemointact perfusions and 130 + 9 sec in the chemodenervated perfusions.

HYPOCAPNIC VERTEBRAL ARTERY PERFUSION

1

.

0

~

,

U

.

103

~INTACT

....

o.oi-

""

F(C02I 0.4 F('n

nI 0.2 O'vo ....

600 ,i

I 0" 0. . . . 2 0| 0 . . . . ~ 0i . . . . 4 0 i0 . . . . 5 0|0 , . ,

TIME (seconds) Fig. 2. The transient respiratory frequency responses (means + SEM) of chemointact and chemodenervated dogs to hypocapnic vertebral artery perfusions at normal flows. Responses are expressed as the ratio to the CO 2 breathing frequency.

I.O i

O.e

O.E

VTm VT(C02)O. 4 O.2 O 0

•0

I

l

l

I

I

I

I

"

I

I

I

I

I

I

I

.

.

.

.

I

I

i

I

i

I

i

I

I

I

I

I00 2OO 300 40O 500 600 TIME (seconds)

Fig. 3. The transient tidal volume responses (mean + SEM), expressed as a ratio to the CO2 breathing tidal volume, of chemointact and chemodenervated dogs to hypocapnic normal flow vertebral artery perfusion. The time courses o f the respiratory frequency changes in chemointact and c h e m o d e n e r v a t e d perfusions were n o t consistently different (fig. 2). The tidal v o l u m e responses, again expressed as a ratio to the C O 2 control tidal volumes, were significantly higher in chemointact than in c h e m o d e n e r v a t e d dogs at 70 sec a n d thereafter (fig. 3). Therefore, differences in ventilatory resp6nses o f the two g r o u p s was primarily due to a maintained higher tidal v o l u m e in the chemointact dogs. In 20 high flow perfusions in 12 chemointact dogs the ventilation, expressed as a ratio to CO2 ventilation, was significantly reduced after 10 sec o f high flow perfusion a n d was significantly lower than the n o r m a l flow perfusions f r o m 40 sec to the conclusion o f the perfusions (fig. 4). After 70 sec o f high flow perfusion the

104

P.J. FEUSTEL et al.

o

"'""'",.

0.8

~/~(T) Vz (c°z)

0.6 0.4

O.2 O0 . . . . . . . . . . • 0 I00

'. . . . . ~ . . . . t .... , ..... 200 300 400 500 600 T I M E (seconds)

Fig. 4. The transient ventilatory responses of intact dogs to normal and high flow perfusions. Room air control VR was 0.21 _+0.02.

~ ,., .

I.O 0.8

CONSTANTPRESSURE

""

"":'~ . . . . . . . .

O.6 FIT) F(C°z) 0.4

O.2

OG . . . . = . . . . i . , ,, = • 0

I00

200 TIME

.

.

.

.

=. . . . i . . . . i

300 400 ( seconds}

500

600

Fig. 5. The transient respiratory frequency responses of chemointact dogs to normal and high flow perfusion.

I.OI .....

VT {C02)VT' {T) 0.80,40.~6

''' '~'"' "'""'" " CONSTANT PRESSURE

i 0.2 O0 i 0

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

I00

200 TIME

300 400 (seconds)

500

600

Fig. 6. The transient tidal volume responses of chemointact dogs to normal and high flow perfusions.

H Y P O C A P N I C VERTEBRAL A R T E R Y P E R F U S I O N

105

TABLE 3 The vertebral artery contribution to regional brain blood flow during hypercapnia and hypocapnic perfusion (ml/min-100 g, Mean _+ SEM) 8°/~ End tidal CO2 (N = 8)

Perfusion constant pressure (N = 9)

Perfusion high flow (N = 9 )

Spinal cord (rostral 2 cm)

R* L

84 + 19 81+_ 19

44 _+ 27 42,+27

96 +_ 43 113+_ 42

Medulla (caudal to obex)

R L

92 +_ 17 143_+ 47

45 _+ 28 59_+36

108 +_ 45 211 + 104

Medulla (rostral to obex)

R L

91 +_ 17 118+_ 19

50 + 27 53+_29

148 _+ 53 172+_ 82

Pons

R L

103 -+ 13 131 + 30

45 + 20 39_+20

141 -+ 67 163_+ 83

R + L R + L

138 +- 13 133 + 20

51 _+ 31 49 _+ 31

144 -+ 65 148 _+ 66

Mesencephalon

R L

128 +- 15 127-+ 14

22 +- 8 19+ 9

95 _+ 33 8 9 + 30

Thalmus

R L

160 +- 49 144+ 46

29 + 17 39+33

156 _+ 90 184+ 91

Hypothalamus

R L

254 +_ 210 306+ 189

16 + 17 10+ 5

64 + 27 6 7 + 35

R L

5 6 + 36 31 +__ 13 3 0 + 14

4+ 2 24 + 17 2 6 + 17

64-+ 41 53 + 19 83_+ 28

R L

30-+ 30+

13 14

23+-18 25+- 17

4 4 + 17 54-+ 22

R L

30+28+

13 15

23+-18 24+16

45+ 44+

R L

33+ 30+

12 13

23+-19 24+- 15

46+_ 17 7 1 + 30

R L

38_+ 13 4 2 + 13

25_+19 25-+17

60_+ 21 49_+ 19

R L

42+ 59+

13 18

26+16 25+_15

63-+ 87+

R L

57+ 46_+

16 12

27+15 2 5 + 19

89-+ 31 8 4 + 36

R L

53 + 19 7 8 + 32

28 _+ 17 26_+17

94 + 36 111_+ 54

Cerebellum (inferior) (superior)

Pituitary Cortical grey and white** Anterior

Posterior

R+L

16 18

18 19

* R = right; L = left. ** The brain was divided midsagittally. Then each half was sectioned coronally at the midpoint of the corpus callosum and at the head and tail of the corpus callosum.

106

P.J. FEUSTEL et al.

VR was 0.69 _+ 0.05 compared with 0.90 _+0.03 during the normal flow perfusions. The faster decrease in ventilation as compared with the normal flow perfusions was due to a faster decline in both frequency and tidal volume (figs. 5 and 6). Mean vertebral artery flow was 65.6 + 16.3 ml/min during CO 2 breathing, 58.6 + 10.6 during hypocapnic normal flow perfusions and 168.2 + 19.6 ml/min during high flow perfusions. The vertebral artery contribution to medullary blood flow was 1.18 _+0.19 ml/min/g of tissue during CO2 breathing, decreased to 0.52 + 0.20 ml/min/g following 3 min of normal flow perfusion but was maintained at 1.05 _+0.12 ml/min/g during high flow perfusion (table 3).

Discussion Previous hypocapnic vertebral artery perfusions with step decreases of 18 to 33 mm Hg during systemic hypercapnia resulted in only a 33~ reduction in ventilation after two minutes of perfusion and failed to reach equilibrium (Dutton et al., 1969). Since Mitchell et al. (1963) and Schlaefke et al. (1970) have shown that hypocapnic or hypercapnic, and hyperhydric or hypohydric, applications of mock CSF to the superficial chemoreceptive tissue of the ventrolateral surface of the medulla results in prompt changes in ventilation, it was reasoned that the time necessary for the clearance of CO2 from a relatively extensive tissue and CSF compartment was responsible for the maintained ventilation. In the present experiments a more rapid decrease in ventilation with the high flow perfusions supports this conclusion. However, the results also suggest that a peripheral chemoreceptor contribution is responsible for at least part of the maintained ventilation after 70 sec of perfusion. In man, using a several breath CO, stimulus, Edelman et al. (1973) observed that transient isoxic hypercapnia results in a ventilatory response which is approximately one third of the steady-state response. In carotid body resected asthmatic patients, Lugliani et al. (1971) found 30~/£ of the response to CO 2 was attributable to the carotid body. Mitchell (1965) has estimated that the peripheral chemoreceptors are responsible for one fourth of the response to CO 2 in anesthetized dogs. By separately perfusing the cat medulla, Heeringa et al. (1979) have attributed 20 to 50g,q of the steady-state hyperoxic CO, response slope to the carotid bodies. The ventilation at the conclusion of vertebral artery hypocapnia in our studies indicates that the peripheral chemoreceptors contribute approximately one fourth of the total increase in ventilation above normocapnic levels. Furthermore, if one assumes that ventilation cannot be sustained in the absence of stimulation to either central or peripheral chemoreceptors and that during hypocapnia the central chemoreceptors make a negligible contribution to ventilatory drive, then in the present study the carotid chemoreceptors are seen to be responsible for one third of the total ventilation in hypercapnia. The former assumption is based on the findings of Loeschcke and Schlafke (1976) in cats that although ventilation was sustained by the peripheral

HYPOCAPNIC VERTEBRALARTERY PERFUSION

107

chemoreceptors after central chemodenervation by coagulation of the central chemosensitive areas, subsequent severing of the carotid sinus nerve and the vagus led to apnea. The latter assumption is based on the present findings of near apnea in 14 of 33 carotid body denervated dogs. Although the ventilatory contribution of the aortic chemoreceptors has been shown to be less than the carotid chemoreceptors in dogs (Daly and Ungar, 1966) and their response to CO: is either small or variable (Fitzgerald, 1976; Paintal and Riley, 1966) in cats, the aortic chemoreceptors may account for the lack of near apnea in the remaining perfusions. The differences in the half times of the responses of chemointact and chemodenervated dogs cannot be explained by any linear combination of two inputs to the respiratory controller. If the central and peripheral stimuli were additive, the half times of the two responses would be the same assuming that the cerebral blood flow responses were the same. Multiplicative interaction of the central and peripheral afferents has been suggested (Loeschcke et al., 1963 ; Cunningham, 1973 ; Adams et al., 1978), but this cannot explain the differences either. Since the peripheral afferent activity is unchanged, the time courses of both the solely central component and the multiplicative component would be expected to be the same and equivalent to the time constant of the central component alone. Although there is evidence to the contrary (Heistad and Marcus, 1978), even if the peripheral chemoreceptors were capable of elevating cerebral blood flow for several minutes when stimulated, as suggested by Ponte and Purves (1974), then the half times of the intact dogs should be less than those of the denervated dogs. Instead, in these experiments, it appears that the peripheral chemoreceptor contribution to ventilation is increasing as the central CO, level is decreasing. The increased peripheral contribution to ventilation cannot be attributed to a developing hypoxia since mean ventilation in the chemointact dogs exceeds room air control ventilation throughout the perfusion. Nor can it be attributed to increasing arterial Pco: since the end-tidal fraction of CO 2 was held at 8~o. These results are consistent with the original findings of Gesell, Lapides and Levin (1940) that peripheral chemoreceptor influence can be removed by cold blocking the carotid sinus nerve without decreasing ventilation during CO 2 breathing. They postulated that the peripheral input is occluded by increased central chemoreceptor activity. Those results are also consistent with the results of Giese et al, (1978), who found that hypoxic stimulation of a separately perfused carotid body had an increasing contribution to total ventilation as central CO~ stimuli decreased. In the present study the chemointact and chemodenervated ventilatory responses differed primarily as a result of a difference in tidal volume responses. The implication is that during hypercapnia the carotid chemoreceptors influence primarily the output of the respiratory centers and do not influence the frequency controller. The observed effect of peripheral chemoreceptor influence primarily on inspiratory output is consistent with the hypothesis of Miserocchi (1976), who found that increasing peripheral afferent activity in hypercapnic hypoxia augmented tidal volumes without altering the tidal volume-inspiratory time relationship. A fixed

108

P.J. FEUSTELet al.

frequency is in contrast with the findings of Rebuck et al. (1976) who found that progressive isocapnic hypoxia increased the frequency of respiration without large changes in tidal volume. When vertebral artery hypocapnic flow was increased in chemointact dogs, ventilation decreased at a faster rate. This suggests that during severe vertebral artery hypocapnia, the lowered blood flow may limit the rate of removal of COz from the central chemoreceptive site. A second possibility is that carotid arterial hypercapnic blood is mixed with vertebral hypocapnic blood during normal flow perfusions. Although this possibility cannot be strictly ruled out, previous studies have shown little or no carotid contribution to medullary or pontine flow (Wellens et al., 1975; Reneman et al., 1974) and holding the vertebral perfusion pressure at approximately five mm Hg above systemic pressure should further insure against mixing. Also, the vertebral artery contribution to medullary flow as measured with microspheres is consistent with measurements made by others of the absolute medullary flow (Marcus et al., 1977; Heistad et al., 1976). Furthermore, the flow to the anterior cerebellum which originates from the anterior cerebellar artery is within the range measured by others (Heistad et al., 1976). This artery originates considerably distal to the proposed chemoreceptor sites. Whether or not cerebral autoregulation is preserved during pump perfusion is questionable (Sagawa and Guyton, 1961 ; Rapela and Green, 1964) and cannot be determined from these data. The cerebral autoregulatory response, if present, would take at least thirty seconds to develop (Rapela and Green, 1964) and therefore would not occur until after a period of high flow sufficient to remove significant quantities of CO, from the chemoreceptor site. A gradual decay of neural activity has been hypothesized to account for the rate of decrease in ventilation following carotid sinus nerve stimulation (Eldridge, 1973) and following CO 2 stimulation of the carotid body (Dutton et al., 1967). The faster decrease in ventilation in the high flow perfusions indicates that the time course of the ventilatory decrease is not governed solely by the rate of decrease of neural activity which was established by preceding chemoreceptor activity. This is not an unexpected result since the time constant of such decays has been reported by Eldridge (1973) to be on the order of 30 sec. It appears more likely that such neural circuits serve to limit the rate of decrease of ventilation after a true step decrease of chemoreceptor input, whereas in these experiments the clearance of CO, is the rate limiting process. In conclusion, it has been shown that the peripheral chemoreceptors are in part responsible for the maintained ventilation during hypocapnic vertebral artery perfusion. The peripheral chemoreceptor contribution to ventilation appears to increase during central hypocapnia. During hypocapnic perfusions a low chemoreceptor blood flow appears to account for the remaining delay in ventilatory recovery from hypercapnia.

HYPOCAPNIC VERTEBRAL ARTERY PERFUSION

109

References Adams, J. M., F.M. Attinger and E.O. Attinger (1978). Medullary and carotid chemoreceptor interaction for mild stimuli. Pfliigers Arch. 374: 39-45. Crone, C. (1965). The permeability of brain capillaries to non-electrolytes. Acta Physiol. Scand. 64: 407-417. Cunningham, D.J.C. (1973). The control system regulating breathing in man. Q. Rev. Biophys. 6: 433-483. Daly, M. De B. and A. Ungar (1966). Comparison of the reflex responses elicited by stimulation of the separately perfused carotid and aortic body chemoreceptors in the dog. J. Physiol. (London) 182: 379-403. Dutton, R.E., V. Chernick, H. Moses, B. Bromberger-Barnea, S. Permutt and R.L. Riley (1964). Ventilatory response to intermittent inspired carbon dioxide. J. Appl. Physiol. 19:931 936. Dutton, R. E., W.A. Hodson, D.G. Davies and V. Chernick (1967). Ventilatory adaptation to a step change in Pco2 at the carotid bodies. J. Appl. Physiol. 23:195 202. Dutton, R.E., D.G. Davies, P.K. Ghatak and R.S. Fitzgerald (1969). Respiration during transient perfusion of vertebral arteries with hypocapnic blood. Am. J. Physiol. 217 : 1178--1182. Edelman, N. H., P. E. Epstein, S. Lahiri and N. S. Cherniack (1973). Ventilatory responses to transient hypoxia and hypercapnia in man. Respir. Physiol. 17:302 314. Eldridge, F.L. (1973). Posthyperventilation breathing: different effects of active and passive hyperventilation. J. Appl. Physiol. 34:422 430. Fitzgerald, R.S., N. Gross and R.E. Dutton (1968). Ventilatory responses to transient acidic and hypercapnic vertebral artery infusions. Respir. Physiol. 4:387 395. Fitzgerald, R.S. (1976). Single fiber chemoreceptor responses of carotid and aortic bodies. In: Morphology and Mechanisms of Chemoreceptors, edited by A.S. Paintal. New Dehli, India, Navchetan Press, pp. 27 35. Gelfand, R. and C.J. Lambertsen (1973). Dynamic respiratory response to abrupt change of inspired CO: at normal and high Po.,. J. Appl. Physiol. 35: 903-913. Gesell, R., J. Lapides and M. Levin (1940). The interaction of central and peripheral chemical control of breathing. Am. J. Physiol. 130: 155-170. Giese, K., J. Berndt and W. Berger (1978). Interaction of central and peripheral respiratory drives in cats. II. Peripheral and central interaction of hypoxia and hypercapnia. Pfli~gers Arch. 374:211-217. Heeringa, J., A. Berkenbosch, J. De Goede and C.N. Olievier (1979). Relative contribution of central and peripheral chemoreceptors to the ventilatory response to CO 2 during hyperoxia. Respir. Physiol. 37:365 379. Heistad, D. D., M. L. Marcus, J. C. Ehrhardt and F. M. Abboud (1976). Effect of stimulation of carotid chemoreceptors on total and regional cerebral blood flow. Circ. Res. 38: 20-25. Heistad, D.D. and M.L. Marcus (1978). Evidence that neural mechanisms do not have important effects on cerebral blood flow. Circ. Res. 42:295 302. Heymans, C. and J. J. Bouckaert (1934). Modifications de la pression art6rielle apr6s section des quatre nerfs freinateurs chez le chien. C.R. Soc. Biol. 117: 252-255. Krasney, J. A. (1971). Cardiovascular responses to cyanide in awake sinoaortic denervated dogs. Am. J. Physiol. 220: 1361-1366. Loeschcke, H.H., R.A. Mitchell, B. Katsaros, J.F. Perkins and A. Konig (1963). Interaction of intracranial chemosensitivity with peripheral afferents to the respiratory centers. Ann. N.Y. Acad. Sci. 109:651 660. Loeschcke, H. H. and M. E. Schlafke (1976). Central chemosensitivity. In: Morphology and Mechanisms of Chemoreceptors, edited by A. S. Paintal. New Dehli, India, Navchetan Press, pp. 282-298. Lugliani, R., B. J. Whipp, C. Seard and K. Wasserman (1971). Effect of bilateral carotid-body resection on ventilatory control at rest and during exercise in man. N. Engl. J. Med. 285:1105 1111.

110

P.J. FEUSTEL et al.

Marcus, M.L., D.D. Heistad, J,C. Ehrhardt and F.M. Abboud (1977). Regulation of total and regional spinal cord blood flow. Circ. Res. 41 : 128-134. Miserocchi, G. (1976). Role of peripheral and central chemosensitive afferents in the control of depth and frequency of breathing. Respir. Physiol. 26:101 111. Mitchell, R.A., H.H. Loeschcke, J.W. Severinghaus, B.W. Richardson and W.H. Massion (1963). Regions of respiratory chemosensitivity on the surface of the medulla. Ann. N.Y. Acad. Sci. 109: 661 681. Mitchell, R.A. (19653. The regulation of respiration in metabolic acidosis and alkalosis. In: Cerebrospinal Fluid and the Regulation of Ventilation, edited by C. McC. Brooks, F.F. Kao and B.B. Lloyd. Philadelphia, PA, F.A. Davis, pp. 109-131. Paintal, A. S. and R. L. Riley (1966). Responses of aortic chemoreceptors. J. Appl. Physiol. 21 : 543-548. Ponte, J. and M.J. Purves (1974). The role of the carotid body chemoreceptors and carotid sinus baroreceptors in the control of cerebral blood vessels. J. Physiol. (London) 237: 315-340. Rapela, C.E. and H.D. Green (1964). Autoregulation of canine cerebral blood flow, Circ. Res. 15 (Suppl. I): 205-212. Rebuck, A. S., J. R.A. Rigg and N.A. Saunders (1976). Respiratory frequency response to progressive isocapnic hypoxia. J. Physiol. (London) 258 : 1%31. Reneman, R.S., D. Wellens, A. H. M. Jageneau and L. Stynen (19743. Vertebral a.nd carotid blood distribution in the brain of the dog and cat. Cardiovasc. Res. 8: 65-72. Rudolph, A. M. and M. A. Heymann (1967). The circulation of the fetus in utero : Method for studying distribution of blood flow, cardiac output and organ blood flow. Circ. Res. 21 : 163 184. Sagawa, K. and A. Guyton (1961). Pressure-flow relationships in isolated canine cerebral circulation. Am. J. Physiol. 200:711 714. Salmoiraghi, G. C. and B. D. Burns (1960). Notes on mechanism of rhythmic respiration. J. Neurophysiol. 23 : 14-26. Schlaefke, M. E., W. R. See, and H. H. Loeschcke (1970). Ventilatory response to alterations of H + ion concentration in small areas of the ventral medullary surface. Respir. Physiol. 10: 198-212. Severinghaus, J.W. and N. Lassen (1967). Step hypocapnia to separate arterial from tissue Pco2 in the regulation of cerebral blood flow. Circ. Res. 20: 272-278. Wellens, D. L. F., L. J. M. R. Wouters, R. J. J. De Reese, P. Beirnaert and R.S. Reneman (1975). The cerebral blood distribution in dogs and cats. An anatomical and functional study. Brain Res. 86: 429438.